Astrophys Space Sci (2009) 320: 77–84DOI 10.1007/s10509-008-9825-4

O R I G I NA L A RT I C L E

Galactic dynamos and galactic winds

Rainer Beck

Received: 26 November 2007 / Accepted: 24 April 2008 / Published online: 30 May 2008© The Author(s) 2008. This article is published with open access at Springerlink.com

Abstract Spiral galaxies host dynamically important mag-netic fields which can affect gas flows in the disks and halos.Total magnetic fields in spiral galaxies are strongest (up to30 µG) in the spiral arms where they are mostly turbulent ortangled. Polarized synchrotron emission shows that the re-solved regular fields are generally strongest in the interarmregions (up to 15 µG). Faraday rotation measures of radiopolarization vectors in the disks of several spiral galaxiesreveal large-scale patterns which are signatures of coherentfields generated by a mean-field dynamo. Magnetic fieldsare also observed in radio halos around edge-on galaxiesat heights of a few kpc above the disk. Cosmic-ray drivengalactic winds transport gas and magnetic fields from thedisk into the halo. The halo scale height and the electron life-time allow to estimate the wind speed. The magnetic energydensity is larger than the thermal energy density, but smallerthan the kinetic energy density of the outflow. There is noobservation yet of a halo with a large-scale coherent dynamopattern. A global wind outflow may prevent the operationof a dynamo in the halo. Halo regions with high degreesof radio polarization at very large distances from the diskare excellent tracers of interaction between galaxies or rampressure of the intergalactic medium. The observed extentof radio halos is limited by energy losses of the cosmic-rayelectrons. Future low-frequency radio telescopes like LO-FAR and the SKA will allow to trace halo outflows and theirinteraction with the intergalactic medium to much larger dis-tances.

R. Beck (�)Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69,53121 Bonn, Germanye-mail: rbeck@mpifr-bonn.mpg.de

Keywords Galaxies: halos · Galaxies: magnetic fields ·Galaxies: spiral · Intergalactic medium · Radio continuum:galaxies

1 Introduction

Galactic halos are dynamical systems receiving their en-ergy from the underlying star-forming disk. Hot thermal gasfrom supernova remnants or superbubbles can drive fountainoutflows, but no steady wind (Breitschwerdt et al. 1991).Cosmic rays accelerated in supernova remnants can pro-vide the pressure to drive a wind. Cosmic rays also drivebuoyant loops of magnetic fields via the Parker instability(Hanasz et al. 2002) (see also Avillez, this volume) and a fastdynamo which amplifies the magnetic field (Parker 1992;Moss et al. 1999; Hanasz et al. 2004). Streaming insta-bilities of the cosmic-ray flow excite plasma waves in thehalo and couple the gas to the outflow (Breitschwerdt et al.1991). Magnetic reconnection can heat the gas (Zimmeret al. 1997). Outflows from starburst galaxies in the earlyUniverse may have magnetized the intergalactic medium(Kronberg et al. 1999), but the role of present-day galaxiesis still unclear. Understanding the interaction between thegas and the magnetic field in the outflow is the key to under-stand the physics of halos and their role during the evolutionof galaxies.

2 The tools of radio synchrotron emission

The intensity of synchrotron emission is a measure of thenumber density of cosmic-ray electrons in the relevant en-ergy range and of the strength of the total magnetic field

78 Astrophys Space Sci (2009) 320: 77–84

component in the sky plane. The degree of linear polariza-tion can be up to 75%. Any variation of the field orientationwithin the beam reduces the degree of polarization. Polar-ized emission emerges from regular fields, which are fieldshaving a constant direction within the telescope beam, orfrom anisotropic magnetic fields, which are generated fromrandom magnetic fields by a compressing or shearing gasflow and frequently reverse their direction by 180° on scalessmaller than the telescope beam. Unpolarized synchrotronemission indicates fields with random directions which havebeen tangled or created by turbulent gas flows.

At short radio wavelengths the orientation of the ob-served polarization vector is perpendicular to the field orien-tation. The orientation of the polarization vectors is changedin a magnetized thermal plasma by Faraday rotation. The ro-tation angle increases with the plasma density, the strengthof the component of the field along the line of sight and thesquare of the observation wavelength. As the rotation an-gle is sensitive to the sign of the field direction, only regu-lar fields can give rise to Faraday rotation, while anisotropicand random fields do not. For typical plasma densities andregular field strengths in the interstellar medium of galaxies,Faraday rotation becomes significant at wavelengths largerthan a few centimeters. Measurements of the Faraday ro-tation from multi-wavelength observations allow to deter-mine the strength and direction of the regular field compo-nent along the line of sight. Its combination with the totalintensity and the polarization vectors can yield the three-dimensional picture of the magnetic field and allows to dis-tinguish the three field components: regular, anisotropic andrandom.

3 Magnetic fields in spiral arms and bars

The total magnetic field strengths can be determined fromthe intensity of the total synchrotron emission, assumingenergy equipartition between the total magnetic field andthe total cosmic rays, and a ratio between the numbersof cosmic-ray protons and electrons in the relevant energyrange (usually �100). The equipartition assumption seemsto be valid in galaxies on large spatial and time scales, butdeviations probably occur on local scales. The typical aver-age equipartition strength of the total magnetic field in spi-ral galaxies is about 10 µG. Radio-faint galaxies like M 31and M 33, our Milky Way’s neighbors, have weaker totalmagnetic fields (about 5 µG), while gas-rich galaxies withhigh star-formation rates, like M 51, M 83 and NGC 6946,have average total field strengths of 15 µG. The mean en-ergy density of the magnetic field and of the cosmic rays inNGC 6946 is �10−11 erg cm−3, about 10 times larger thanthat of the ionized gas, but similar to that of the turbulent gasmotions across the whole star-forming disk (Beck 2007a).

The magnetic energy may even dominate in the outer disk.The strongest fields (50–100 µG) are found in starburstgalaxies, like M 82 and the “Antennae” NGC 4038/9, andin nuclear starburst regions, like in the centers of NGC 1097and other barred galaxies (Beck 2005).

Spiral arms in total radio emission appear very similarto those observed in the far-infrared. The total equipartitionfield strength in the arms can be up to 30 µG. The degreeof radio polarization within the spiral arms is only a few %;hence the field in the spiral arms must be mostly tangledor random within the telescope beam, which typically cor-responds to a few 100 pc. Strong random fields in spiralarms are probably generated by turbulent gas motions or theturbulent dynamo (Brandenburg and Subramanian 2005). Incontrast, the ordered (regular or anisotropic) fields traced bythe polarized synchrotron emission are generally strongest(10–15 µG) in the regions between the optical spiral arms,oriented parallel to the adjacent optical spiral arms. In sev-eral galaxies the field forms “magnetic arms” between theoptical arms, like in NGC 6946 (Beck 2007a). These areprobably generated by the mean-field dynamo (Sect. 4). Ingalaxies with strong density waves some of the ordered fieldis concentrated at the inner edge of the spiral arms, e.g. inM 51 (Patrikeev et al. 2006).

The ordered magnetic field forms spiral patterns in al-most every galaxy, even in flocculent and bright irregularones which lack any optical spiral structure (Beck 2005),a strong argument for dynamo action. Spiral fields are alsoobserved in the central regions of galaxies and in circum-nuclear gas rings.

In galaxies with massive bars, the field seems to followthe gas flow. As there (inside the corotation radius) the gasrotates faster than the spiral or bar pattern of a galaxy, ashock occurs in the cold gas which has a small sound speed,while the warm, diffuse gas is only slightly compressed. Asthe observed compression of the field in spiral arms and barsis also small, the ordered field is coupled to the warm gasand is strong enough to affect the flow of the warm gas (Becket al. 2005). The polarization pattern is an excellent tracer ofthe gas flow in the sky plane and hence complements spec-troscopic measurements.

4 Galactic dynamos

The origin of the first magnetic fields in the Universe is stilla mystery (Widrow 2002). Protogalaxies probably were al-ready magnetic due to field ejection from the first stars orfrom jets generated by the first black holes. A large-scaleprimordial field in a young galaxy is hard to maintain be-cause the galaxy rotates differentially, so that field lines getstrongly wound up during galaxy evolution, in contrast tothe observations which show large pitch angles. This calls

Astrophys Space Sci (2009) 320: 77–84 79

Fig. 1 Dynamo field of type A0: The toroidal field component (thicklines) is axisymmetric in the plane (azimuthal mode m = 0) and re-verses its sign across the plane. The poloidal field component (thinlines) is of dipole type (Han 2002)

for a mechanism to sustain and organize the magnetic field.The most promising mechanism is the dynamo (Beck et al.1996; Brandenburg and Subramanian 2005) which generateslarge-scale regular fields, even if the seed field was turbulent.

The regular field structure obtained in models of themean-field α�-dynamo, driven by turbulent gas motionsand differential rotation, is described by modes of differentazimuthal symmetry in the disk and vertical symmetry per-pendicular to the disk plane. Several modes can be excitedin the same object.

In spherical bodies like stars, planets or galaxy halos,the strongest mode is an antisymmetric toroidal field witha sign reversal across the equatorial plane, surrounded bya poloidal dipole field which is continuous across the plane(mode A0, Fig. 1). For thick galaxy disks or halos the S0 andA0 modes of strong dynamos are oscillatory, i.e. the fieldtorus migrates radially, so that large-scale reversals in thedisk at certain radii are expected (Elstner et al. 1992).

In flat objects like galaxy disks, the strongest mode (S0)is a toroidal field which is symmetric with respect to theplane and has the azimuthal symmetry of an axisymmetricspiral in the plane, without sign reversals, surrounded by aweaker poloidal field of quadrupole structure with a reversalof the vertical field component across the equatorial plane.The next strongest mode is of bisymmetric spiral shape (S1)with two sign reversals within the disk, followed by morecomplicated modes (Baryshnikova et al. 1987).

The standard mean-field α�-dynamo in galaxy disks am-plifies the field and builds up large-scale coherent fieldswithin about 109 yr (Beck et al. 1994b). Faster amplificationis possible when cosmic-ray driven Parker loops and fieldreconnection are involved (Parker 1992; Hanasz and Lesch1998; Moss et al. 1999; Hanasz et al. 2004). The small-scale or fluctuation dynamo (Brandenburg and Subramanian2005) amplifies turbulent, incoherent magnetic fields within106–107 yr.

Kinematical dynamo models including the velocity fieldof a global galactic wind show field structures which are

Fig. 2 Dynamo model for the galaxy NGC 891 with a galactic wind:predicted contours of polarized radio emission and vectors (left) andfield geometry (right) for a radial wind with a terminal velocity of (a) 0,(b) 50 and (c) 200 km/s. The solid and dashed lines indicate differentfield directions; the field is oscillatory and reverses its direction alongradius and height (from Brandenburg et al. 1993, Fig. 9)

parallel to the plane in the inner disk, but open radially out-wards, depending on the wind speed and direction (Branden-burg et al. 1993). The model for the galaxy NGC 891 showsan oscillating field (Fig. 2), while for other galaxy modelsa slow wind speed of 50 km/s was found to be sufficient tochange the oscillating mode into a steady one.

For fast global winds the advection time for the field maybecome smaller than the dynamo amplification time, so thatthe dynamo cannot operate and the field gas becomes frozeninto the flow. Non-uniform local (“spiky”) winds allow dy-namo action, but modify the field configuration into oscillat-ing dipoles or quadrupoles with dominating vertical compo-nents and frequent, migrating reversals (Elstner et al. 1995).However, strong fields may back-react onto the wind. Dy-namical models are needed, including the interplay betweenthe gas flow and the magnetic field.

80 Astrophys Space Sci (2009) 320: 77–84

Dynamo modes can be identified from the pattern of po-larization angles and Faraday rotation measures (RM) inmulti-wavelength radio observations of galaxy disks withmoderate inclination (Elstner et al. 1992; Krause 1990) orfrom RM data of polarized background sources (Stepanovet al. 2008). The disks of a few spiral galaxies indeed re-veal large-scale RM patterns, as predicted. The Andromedagalaxy M 31 hosts a dominating axisymmetric disk field,the basic azimuthal dynamo mode S0 (Fletcher et al. 2004),which extends to at least 15 kpc distance from the center(Han et al. 1998). Other candidates for a dominating axisym-metric disk field are the nearby spiral IC 342 (Krause et al.1989) and the irregular Large Magellanic Cloud (LMC)(Gaensler et al. 2005). The magnetic arms in NGC 6946 canbe described by a superposition of two azimuthal dynamomodes, where the dynamo wave is phase shifted with re-spect to the density wave (Beck 2007a). However, in manyobserved galaxy disks no clear patterns of Faraday rotationwere found. Either several dynamo modes are superimposedand cannot be distinguished with the limited sensitivity andresolution of present-day telescopes, or no large-scale dy-namo modes exist and most of the ordered fields traced bythe polarization vectors are anisotropic (with frequent rever-sals), due to shearing or compressing gas flows.

5 Radio halos of edge-on galaxies

Radio halos are observed around the disks of many edge-ongalaxies (Hummel et al. 1991b; Irwin et al. 1999), but theirradio intensity and extent varies significantly. The radial ex-tent of radio halos increases with the size of the activelystar-forming parts of the galaxy disk (Dahlem et al. 2006).The halo luminosity in the radio range correlates with thosein Hα and X-rays (Tüllmann et al. 2006), although the de-tailed halo shapes vary strongly between the different spec-tral ranges. These results suggest that star formation in thedisk is the energy source for halo formation. Dahlem et al.(1995) argue that the energy input from supernova explo-sions per surface area in the projected disk determines thehalo size.

In spite of their largely different intensities and ex-tents, the scale heights of radio halos observed at 5 GHzare �1.8 kpc (Dumke and Krause 1998; Krause 2004)and surprisingly similar. Their sample of galaxies includedone of the weakest halos, NGC 4565 (Fig. 3) as well asone of the brightest, NGC 891 (Fig. 4). Assuming en-ergy density equipartition between magnetic fields and cos-mic rays, the scale height of the total magnetic field isabout 4 times larger than that of the synchrotron emis-sion, hence �7 kpc on average. A prominent exceptionis NGC 4631 with the largest radio halo (scale height of�2.5 kpc) observed so far (Hummel and Dettmar 1990;

Fig. 3 Optical image of the edge-on spiral galaxy NGC 4565, overlaidby contours of the intensity of the total radio emission at 6 cm wave-length and polarization vectors, combined from observations with theVLA and the Effelsberg 100-m telescope (Dumke 1997). The field linesare mostly parallel to the disk. (Copyright: MPIfR Bonn)

Fig. 4 Optical image of the edge-on spiral galaxy NGC 891, overlaidby contours of the intensity of the total radio emission at 3.6 cm wave-length and polarization vectors, observed with the Effelsberg 100-mtelescope (Krause 2007). The field lines are parallel to the disk near theplane, but turn vertically above and below the disk. (Copyright: MPIfRBonn)

Hummel et al. 1991a; Krause 2004) (Fig. 7). In case of en-ergy equipartition, the scale height of the total field is atleast (3 + α) times larger than the synchrotron scale height

Astrophys Space Sci (2009) 320: 77–84 81

(where α is the synchrotron spectral index), hence ≥10 kpcin NGC 4631.

The scale height may be larger if cosmic rays origi-nate from star-forming regions in the plane and are not re-accelerated in the halo, so that the electrons lose their energyabove some height and the equipartition formula yields toosmall values for the field strength (Beck and Krause 2005).As the degree of linear polarization increases with heightabove the disk midplane, the scale height of the ordered fieldmay be even larger.

Due to its large scale height, the magnetic energy den-sity in halos is much larger than that of the thermal gas(Ehle et al. 1998), while still smaller than the dominatingkinetic energy of the wind outflow. Hence the magnetic fieldis frozen into the outflow and its structure is a signature ofthe outflow kinematics (Sect. 6).

Radio halos grow with decreasing observation frequency,which indicates that the extent is limited by energy losses ofthe cosmic-ray electrons, i.e. synchrotron, inverse Compton,bremsstrahlung and adiabatic losses (Pohl and Schlickeiser1990). The stronger magnetic field in the central regionsleads to larger synchrotron loss, leading to the squeezedshape of many halos, e.g. NGC 253 (Carilli et al. 1992;Beck et al. 1994a; Heesen et al. 2005) (Fig. 6), which isin contrast to its almost spherical X-ray halo (Pietsch et al.2000). On the other hand, the exceptionally large and almostspherical radio halo above the central region of NGC 4631(Hummel and Dettmar 1990) (see also Fig. 7) indicates thatthe magnetic field is weaker and/or the galactic wind is fasterthan in other galaxies.

The two different transport mechanisms for cosmic rays,diffusion or advection (galactic wind), can be distinguishedand modeled from the variation of the radio spectral in-dex with distance from the plane, to be determined frommulti-frequency radio observations (Pohl and Schlickeiser1990). Breaks in the spectra of total radio emission due togalactic winds were detected in several radio halos (Pohlet al. 1991a). NGC 4631 is also exceptional concerning itsradio spectrum (Pohl et al. 1991b). Heesen et al. (2008)measured a transport speed of about 200 km/s in the haloNGC 253 from the radio scale height and the synchrotronloss time.

Note that most of the large-scale halo emission is miss-ing in some VLA maps (Sukumar and Allen 1991; Carilliet al. 1992; Soida 2005), which also severely affects the po-larization angles. A combination with single-dish (Effels-berg) data is indispensable for large galaxies and at wave-lengths ≤ 6 cm (Figs. 3 and 6), while at ≥ 20 cm the VLA issensitive to extended structures up to ≥ 10′, which is suf-ficient to detect the halos of the largest edge-on galaxies(Fig. 7).

6 Magnetic field structure in halos

Radio polarization observations of nearby galaxies seenedge-on generally show a disk-parallel field near the diskplane (Dumke et al. 1995). High-sensitivity observations ofseveral edge-on galaxies like NGC 891 (Fig. 4), NGC 5775(Fig. 5), NGC 253 (Fig. 6) and M 104 (Krause et al. 2006)show vertical field components which increase with increas-ing height z above and below the galactic plane and alsowith increasing radius, so-called X-shaped magnetic fields(Krause 2007).

The observation of X-shaped polarization patterns is offundamental importance to understand the origin of mag-netic fields in halos. The X-shape is inconsistent with thepredictions from standard dynamo models without windflows (Sect. 4), so that the field has to be transported fromthe disk into the halo. A uniform wind would just lift up thedisk field and can also be excluded. A superwind emergingfrom a starburst nucleus, as predicted for NGC 253 (Heck-man et al. 1990), would produce a radial pattern, whichis not observed. The most probable explanation is a windemerging from the disk which is modified by differential ro-tation in the halo and dynamo action.

The analysis of the highly inclined galaxy NGC 253(Fig. 6) allowed a separation of the observed field into an ax-isymmetric disk field and a halo field inclined by about 50°(Heesen et al. 2008). Similar tilt angles are also observed atlarge heights in other edge-on galaxies.

Fig. 5 Hα image of the edge-on spiral galaxy NGC 5775, overlaidby contours of the intensity of the total radio emission at 6 cm wave-length and polarization vectors, observed with the VLA (Tüllmannet al. 2000). The field lines are parallel to the disk near the plane, butbecome increasingly vertical away from the plane. (Copyright: CracowObservatory)

82 Astrophys Space Sci (2009) 320: 77–84

Fig. 6 Total radio emission at 6 cm wavelength and polarization vec-tors of the almost edge-on spiral galaxy NGC 253, combined from ob-servations with the VLA and the Effelsberg 100-m telescope (Heesenet al. 2005). (Copyright: AIRUB Bochum/MPIfR Bonn)

Fig. 7 Total radio emission at 22 cm wavelength of the edge-on spiralgalaxy NGC 4631, observed with the VLA (Beck 2005). The polariza-tion vectors are corrected for Faraday rotation. The lack of polarizedemission near the plane is due to Faraday depolarization. (Copyright:MPIfR Bonn)

The field lines in the outer halo of NGC 4631 suggesta dipole field (Fig. 7), but Faraday rotation measures donot support this interpretation (see Sect. 7). The radio haloabove the inner disk is composed of almost vertical magneticspurs connected to star-forming regions (Golla and Hum-mel 1994). The observations support the idea of a strong“spiky” galactic wind which is driven by regions of star for-mation in the inner disk (Elstner et al. 1995). Furthermore,differential rotation is small in the inner disk, so that ver-

tical field lines are less twisted. At larger radii, where starformation is weaker and differential rotation is larger, thefield becomes parallel to the disk. Another galaxy with ver-tical fields away from the plane and hence a strong wind isNGC 4666 (Dahlem et al. 1997; Soida 2005).

7 Faraday rotation and depolarization in halos

Polarization “vectors” are ambiguous by 180° and hencecannot distinguish between a halo field which is sheared intoelongated Parker loops or a regular halo field generated bya dynamo. A large-scale regular field can be measured onlyby Faraday rotation measures (RM). In external galaxies, thevertical field symmetry could not yet been determined fromexisting RM data with sufficient accuracy. RM values in ha-los, e.g. in NGC 253 (Beck et al. 1994a) and in NGC 4631,are small and do not show large-scale patterns. Indirect ev-idence for preferred symmetric toroidal fields follows fromthe possible dominance of the inward-directed radial fieldcomponent in the disk (Krause and Beck 1998). Such a dom-inance of is in conflict with a reversal of the toroidal fieldacross the plane1 in which case the observed field directionwould depend on the aspect angle and no preference wouldbe expected for a galaxy sample.

In the Milky Way, the pattern of rotation measures frompulsars and polarized background sources indicates a sym-metric local magnetic field, but an antisymmetric field inthe halo, i.e. the toroidal field reverses its sign across theplane (dynamo mode A0, Fig. 1) (Han et al. 1997; Han 2002;Sun et al. 2008).

Faraday depolarization is another method to detect mag-netic fields and ionized gas in galaxy halos. In NGC 891and NGC 4631 the mean degree of polarization at 1.4 GHzincreases from about 1% in the plane to about 20% in the up-per halo. This was modeled by depolarization due to randommagnetic fields of 10 µG and 7 µG strength, respectively,and ionized gas with densities in the plane of 0.03 cm−3 and0.07 cm−3 and scale heights of 0.9 kpc and 1.3 kpc, respec-tively (Hummel et al. 1991a).

In face-on galaxies, structures in the halo may becomevisible via depolarization of the underlying disk emission,e.g. in M 51 (Berkhuijsen et al. 1997). A large depolarizedregion above the disk of NGC 6946 may be due to a strongvertical field (Beck 2007a).

1Note that the observed Faraday rotation measure RM is not zero alonga line of sight passing through a disk containing a field reversal in itsmidplane if cosmic-ray electrons and thermal plasma are mixed in thedisk. In this case RM is half of that from a disk without a reversal, andthe sign of RM traces the field direction in the layer which is nearer tothe observer.

Astrophys Space Sci (2009) 320: 77–84 83

8 Interactions

Interaction between galaxies or with a dense intergalacticmedium imprints unique signatures onto magnetic fields ingalaxy halos and thus onto the radio emission. The Virgocluster is a location of strong interaction effects. Highlyasymmetric distributions of the polarized emission showsthat the magnetic fields of several spirals are strongly com-pressed on one side of the galaxy (Vollmer et al. 2007;Wezgowiec et al. 2007).

Interaction may also induce violent star-formation activ-ity in the nuclear region or in the disk which may producehuge radio lobes due to outflowing gas and magnetic field.The lobes of the Virgo spiral NGC 4569 reach out to at least25 kpc from the disk and are highly polarized (Chyzy et al.2006). However, there is no indication for a recent starburst,so that the radio lobes are probably a signature of activity inthe past.

Polarized radio emission is an excellent observationaltracer of interactions. As the decompression timescale of thefield is very long, it keeps memory of events in the past.These are still observable if the lifetime of the illuminat-ing cosmic-ray electrons is sufficiently large. Radio obser-vations at low frequencies are preferable.

9 Summary and outlook

Total and polarized radio emission of edge-on galaxies isa powerful tool to study the physics of galaxy halos. Theextent and spectral index of radio halos gives informationon the transport of cosmic-ray electrons and their originin star-forming regions in the disk. Estimates of the elec-tron lifetime allow an estimate of the transport speed. Theshape of the halo and the field structure may allow to dis-tinguish a global wind driven by star formation in the diskfrom a superwind emerging from a central starburst. Thefield structure in the halo reflects the interaction between theoutflow and differential rotation. Shock fronts between col-liding halo outflows may re-accelerate cosmic rays and en-hance the radio emission. Halo fields compressed by galaxyinteractions or ram pressure with the intergalactic gas are ob-servable via polarized emission. Faraday rotation measurestrace regular fields and can test dynamo models (Stepanovet al. 2008). Faraday rotation and depolarization are alsosensitive tools to detect ionized gas in galaxies and in theintergalactic space.

Future radio telescopes will widen the range of observ-able phenomena around galaxies. High-resolution observa-tions at high frequencies with the Extended Very Large Ar-ray (EVLA) and the planned Square Kilometre Array (SKA)are required to show whether the halo fields are regularor composed of many stretched loops. Forthcoming low-frequency radio telescopes like the Low Frequency Array

(LOFAR), Murchison Widefield Array (MWA), Long Wave-length Array (LWA) and the low-frequency SKA will besuitable instruments to search for extended synchrotron radi-ation at the lowest possible levels in the outer galaxy halos,the transition to intergalactic space and in galaxy clustersand will give access to the so far unexplored domain of weakmagnetic fields in galaxy halos (Beck 2007b, 2008). The de-tection of radio emission from the intergalactic medium willallow to probe the existence of magnetic fields in such rar-ified regions, measure their intensity, and investigate theirorigin and their relation to the structure formation in theearly Universe.

As Faraday rotation angles increase with λ2, low-frequency telescopes will also be able to measure very smallFaraday rotation measures (RM) and Faraday depolariza-tion, which allow to detect weak magnetic fields and low-density ionized gas in galaxy halos. The present limit inelectron densities of about 0.003 cm−3 can be reduced by atleast one order of magnitude.

The planned all-sky RM survey with the SKA at about1 GHz will be used to model the structure and strength ofthe magnetic fields in the intergalactic medium, the inter-stellar medium of intervening galaxies and in the MilkyWay (Beck and Gaensler 2004). With deep polarization andRM observations the origin and evolution of galactic mag-netic fields will be investigated. “Cosmic magnetism” hasbeen accepted as Key Science Projects for both, LOFAR andSKA.

Open Access This article is distributed under the terms of the Cre-ative Commons Attribution Noncommercial License which permitsany noncommercial use, distribution, and reproduction in any medium,provided the original author(s) and source are credited.

References

Baryshnikova, I., Shukurov, A., Ruzmaikin, A., Sokoloff, D.D.: As-tron. Astrophys. 177, 27 (1987)

Beck, R.: In: Wielebinski, R., Beck, R. (eds.) Cosmic Magnetic Fields.Lecture Notes in Physics, vol. 664, p. 41. Springer, Berlin (2005)

Beck, R.: Astron. Astrophys. 470, 539 (2007a)Beck, R.: Adv. Radio Sci. 5, 399 (2007b)Beck, R.: Rev. Mex. Astron. Astrophys. (2008, in press),

arxiv:0804.4594Beck, R., Gaensler, B.M.: New Astr. Rev. 48, 1289 (2004)Beck, R., Krause, M.: Astron. Nachr. 326, 414 (2005)Beck, R., Carilli, C.L., Holdaway, M.A., Klein, U.: Astron. Astrophys.

292, 409 (1994a)Beck, R., Poezd, A.D., Shukurov, A., Sokoloff, D.D.: Astron. Astro-

phys. 289, 94 (1994b)Beck, R., Brandenburg, A., Moss, D., Shukurov, A., Sokoloff, D.:

Annu. Rev. Astron. Astrophys. 34, 155 (1996)Beck, R., Fletcher, A., Shukurov, A., et al.: Astron. Astrophys. 444,

739 (2005)Berkhuijsen, E.M., Horellou, C., Krause, M., et al.: Astron. Astrophys.

318, 700 (1997)Brandenburg, A., Subramanian, K.: Phys. Rep. 417, 1 (2005)

84 Astrophys Space Sci (2009) 320: 77–84

Brandenburg, A., Donner, K.J., Moss, D., et al.: Astron. Astrophys.271, 36 (1993)

Breitschwerdt, D., McKenzie, J.F., Völk, H.J.: Astron. Astrophys. 245,79 (1991)

Carilli, C.L., Holdaway, M.A., Ho, P.T.P., de Pree, C.G.: Astrophys. J.399, L59 (1992)

Chyzy, K.T., Soida, M., Bomans, D.J., et al.: Astron. Astrophys. 447,465 (2006)

Dahlem, M., Lisenfeld, U., Golla, G.: Astrophys. J. 444, 119 (1995)Dahlem, M., Petr, M.G., Lehnert, M.D., Heckman, T.M., Ehle, M.: As-

tron. Astrophys. 320, 731 (1997)Dahlem, M., Lisenfeld, U., Rossa, J.: Astron. Astrophys. 457, 121

(2006)Dumke, M.: 1997, PhD thesis, University of BonnDumke, M., Krause, M.: In: Breitschwerdt, D., Freyberg, M.J., Trüm-

per, J. (eds.) The Local Bubble and Beyond. Lecture Notes inPhysics, vol. 506, p. 555. Springer, Berlin (1998)

Dumke, M., Krause, M., Wielebinski, R., Klein, U.: Astron. Astrophys.302, 691 (1995)

Ehle, M., Pietsch, W., Beck, R., Klein, U.: Astron. Astrophys. 329, 39(1998)

Elstner, D., Meinel, R., Beck, R.: Astron. Astrophys. Suppl. Ser. 94,587 (1992)

Elstner, D., Golla, G., Rüdiger, G., Wielebinski, R.: Astron. Astrophys.297, 77 (1995)

Fletcher, A., Berkhuijsen, E.M., Beck, R., Shukurov, A.: Astron. As-trophys. 414, 53 (2004)

Gaensler, B.M., Haverkorn, M., Staveley-Smith, L., et al.: Science 307,1610 (2005)

Golla, G., Hummel, E.: Astron. Astrophys. 284, 777 (1994)Han, J.: In: Cecchini, S., Cortiglioni, S., Sault, R., Sbarra, C. (eds.)

Astrophysical Polarized Backgrounds. Am. Inst. of Phys. Conf.Ser., vol. 609, p. 96 (2002)

Han, J.L., Manchester, R.N., Berkhuijsen, E.M., Beck, R.: Astron. As-trophys. 322, 98 (1997)

Han, J.L., Beck, R., Berkhuijsen, E.M.: Astron. Astrophys. 335, 1117(1998)

Hanasz, M., Lesch, H.: Astron. Astrophys. 332, 77 (1998)Hanasz, M., Otmianowska-Mazur, K., Lesch, H.: Astron. Astrophys.

386, 347 (2002)Hanasz, M., Kowal, G., Otmianowska-Mazur, K., Lesch, H.: Astro-

phys. J. 605, L33 (2004)Heckman, T.M., Armus, L., Miley, G.K.: Astrophys. J. Suppl. Ser. 74,

833 (1990)Heesen, V., Krause, M., Beck, R., Dettmar, R.-J.: In: Chyzy, K.T.,

Otmianowska-Mazur, K., Soida, M., Dettmar, R.-J., (eds.) TheMagnetized Plasma in Galaxy Evolution, p. 156. JagiellonianUniversity, Kraków (2005)

Heesen, V., Dettmar, R.-J., Krause, M., Beck, R.: Astron. Astrophys.(2008) arxiv:0801.3542

Hummel, E., Dettmar, R.-J.: Astron. Astrophys. 236, 33 (1990)Hummel, E., Beck, R., Dahlem, M.: Astron. Astrophys. 248, 23

(1991a)Hummel, E., Beck, R., Dettmar, R.-J.: Astron. Astrophys. Suppl. Ser.

87, 309 (1991b)Irwin, J.A., English, J., Sorathia, B.: Astron. J. 117, 2102 (1999)Krause, M.: In: Beck, R., Wielebinski, R., Kronberg, P.P. (eds.) Galac-

tic and Intergalactic Magnetic Fields. IAU Symposium, vol. 140,p. 187. Kluwer Academic, Dordrecht (1990)

Krause, M.: In: Uyaniker, B., Reich, W., Wielebinski, R., (eds.) TheMagnetized Interstellar Medium, p. 173. Copernicus, Katlenburg(2004)

Krause, M.: Mem. Soc. Astron. Ital. 78, 314 (2007)Krause, F., Beck, R.: Astron. Astrophys. 335, 789 (1998)Krause, M., Hummel, E., Beck, R.: Astron. Astrophys. 217, 4 (1989)Krause, M., Wielebinski, R., Dumke, M.: Astron. Astrophys. 448, 133

(2006)Kronberg, P.P., Lesch, H., Hopp, U.: Astrophys. J. 511, 56 (1999)Moss, D., Shukurov, A., Sokoloff, D.: Astron. Astrophys. 343, 120

(1999)Parker, E.N.: Astrophys. J. 401, 137 (1992)Patrikeev, I., Fletcher, A., Stepanov, R., et al.: Astron. Astrophys. 458,

441 (2006)Pietsch, W., Vogler, A., Klein, U., Zinnecker, H.: Astron. Astrophys.

360, 24 (2000)Pohl, M., Schlickeiser, R.: Astron. Astrophys. 234, 147 (1990)Pohl, M., Schlickeiser, R., Hummel, E.: Astron. Astrophys. 250, 302

(1991a)Pohl, M., Schlickeiser, R., Lesch, H.: Astron. Astrophys. 252, 493

(1991b)Soida, M.: In: Chyzy, K.T., Otmianowska-Mazur, K., Soida, M.

Dettmar, R.-J. (eds.) The Magnetized Plasma in Galaxy Evolu-tion, p. 185. Jagiellonian University, Kraków (2005)

Stepanov, R., Arshakian, T.G., Beck, R., Frick, P., Krause, M.: Astron.Astrophys. 480, 45 (2008)

Sukumar, S., Allen, R.J.: Astrophys. J. 382, 100 (1991)Sun, X.H., Reich, W., Waelkens, A., Enßlin, T.A.: Astron. Astrophys.

477, 573 (2008)Tüllmann, R., Dettmar, R.-J., Soida, M., Urbanik, M., Rossa, J.: As-

tron. Astrophys. 364, L36 (2000)Tüllmann, R., Breitschwerdt, D., Rossa, J., Pietsch, W., Dettmar, R.-J.:

Astron. Astrophys. 457, 779 (2006)Vollmer, B., Soida, M., Beck, R., et al.: Astron. Astrophys. 464, L37

(2007)Wezgowiec, M., Urbanik, M., Vollmer, B., et al.: Astron. Astrophys.

471, 93 (2007)Widrow, L.M.: Rev. Mod. Phys. 74, 775 (2002)Zimmer, F., Lesch, H., Birk, G.T.: Astron. Astrophys. 320, 746 (1997)